The present invention relates to managing a reboot of a computing node, and more particularly to determining a reboot time of a computing node in a cloud computing environment.
As more cloud environments emerge, cloud security becomes an important issue. In the realm of cloud security, auto patching at operating system (OS) and application levels and critical changes to remediate security vulnerabilities require an OS reboot of cloud nodes. Known techniques of cloud node rebooting include a time-consuming requirement that cloud providers check with cloud virtual machine (VM) owners and receive user input before rebooting. Other known cloud node rebooting techniques include rebooting VMs without notification or pre-alert, which results in application breach or a loss of data, which eventually damages the quality of cloud services.
Factors that affect OS rebooting of cloud nodes are significant and complicated. These factors include determining whether a user is active or inactive, whether an application, file, or process is active or inactive, whether there is live connectivity to target machines. A determination must be made whether these factors are equally important or whether one or more factors are more important than other factor(s) when making a decision to reboot an OS of a cloud node.